Lung cancer causes the deaths of more men and women in the United States than the next four most common types of cancer combined. Only a small fraction of cases are cured, and median survival is short. Some patients will die of metastatic disease, and some with only local disease. Some will respond to chemotherapy and some will not. While the molecular biology of lung cancer has been extensively studied, these variable outcomes cannot be predicted (or understood at a molecular level) by histopathology or currently available molecular markers. In this project, we intend to collect lung cancer tumor specimens from a national cooperative group clinical trial and utilize the unique combination of both protein and RNA-based technologies available at Vanderbilt to develop comprehensive molecular fingerprints of lung cancer. The goal is to determine molecular patterns that will be more predictive of tumor behavior, and may ultimately lead us to a better understanding of this behavior, improved interventions, and improved outcomes. The protein technology to be applied to this question utilizes novel microscopic laser-directed protein mass spectrometric analysis of tumor samples after laser capture microdissection. The same tumor samples will be used to probe large cDNA expression arrays, and the data from each analyzed for statistically significant signatures in defined sets of tumors with complete clinical follow-up. Analytical and statistical techniques will be developed as needed to analyze the data derived from these two technologies. Specifically, we will address Stage I node negative resectable non-small cell lung cancer (NSCLC) patients. In these samples we will identify molecular signatures of the following biologic behaviors: 1) occult lymph nodal involvement at the time of thoracotomy, 2) recurrence with metastatic disease vs. local recurrence, and 3) disease free and overall survival. This unique combination of clinical, technical, and analytical resources will be combined in this proposal to discern and evaluate molecular fingerprints of biology in lung cancer and these data may then lead to increased understanding of the basis of this biology and improved outcomes for patients with lung cancer.